PAC Dosing System Design Criteria: 2026 Engineering Specs

What "PAC Dosing System Design Criteria" Actually Means in 2026

PAC dosing system design criteria is the engineering checklist that turns a bench-scale jar-test dose into a working industrial skid: feed characterization, dose determination, storage, metering, and closed-loop control. In 2026 a defensible specification ties each of those five items to a published code and a measurable number. The codes a buyer should see in the datasheet are ASME B73.1 for chemical process pumps, ANSI/HI 9.6.6 for reciprocating positive-displacement metering-pump turndown, API 650 and NFPA 30 for atmospheric storage of liquid coagulant, and ISA-88 batch-control concepts for the PLC sequence. Anything less and the skid is being specified by analogy to a swimming-pool feeder, which is the mistake most top-ranked pages still make.

The pool-vs-plant gap is not cosmetic. PWTAG's technical note (S3) recommends roughly 0.1 ml/m³ of circulation flow, dosed from a peristaltic pump directly out of the drum, with at least 10 seconds of contact at ≤1.5 m/s velocity. That is correct for a 225 m³/h pool. An industrial clarifier handling 5–500 m³/h of variable raw water cannot be run on the same hardware: flow swings of 2–3× are normal, alkalinity drifts, and the dose that worked on Monday restabilizes colloids on Friday. The skid has to absorb that variability, which is why the 2026 criteria below lean on PLC-based flow pacing plus streaming-current trim, not on a drum and a squeeze-tube pump.

Get the design wrong in either direction and the operator pays for it. Overdosing reverses the colloidal charge, producing a cloudy supernatant that defeats the clarifier (HydroChemix, S2, 2026-08). Underdosing passes turbidity downstream and shortens filter runtimes. The rest of this article builds the five-criteria specification a process engineer can hand to procurement.

Step 1 — Establish the Target Dose with Jar Testing

Jar testing is the only method that ties dose to a specific raw water, and it is the setpoint the PLC will eventually multiply by flow. The bench procedure is standardized: 1 L beakers, six to eight dose points bracketing the expected band, rapid mix at 120–150 rpm for 1 min, slow mix at 30 rpm for 20 min, settle 30 min, then measure supernatant turbidity, residual color, and residual aluminium. Use a jar test procedure that is repeatable across operators — the dose is only as defensible as the bench technique behind it.

For most industrial wastewaters the expected band is 5–150 mg/L of commercial PAC liquid (RubMech, S4); in low-alkalinity drinking water a 2025 GA-optimized full-scale study narrowed the practical range to 4–12 ppm (MDPI, S5, 2025-08). pH should sit in the 5.5–8.5 window for the dose to be effective (S4). Decision rule: pick the lowest dose that clears the turbidity/color target without leaving a positive streaming-current reading at the outlet. Overdosing symptoms show up first as sticky flocs, floating sludge, and a milky supernatant (HydroChemix, S2).

The output of jar testing is a single number — mg/L — that becomes the dosing setpoint. The PLC will multiply that number by measured flow to command the pump, so any error in the bench result propagates linearly into chemical spend. Spend the day on it.

Step 2 — Size the Storage and Day Tanks

Storage tank sizing for an industrial polyaluminum chloride dosing skid is a 7–14 day consumption problem, not a "fits in the corner" problem. Bulk PAC is delivered as a 10–17% Al₂O₃ liquid (the MDPI study specifies 10% w/w as common, S5), and the working formula is:

Tank volume (m³) = [Dose (mg/L) × Flow (m³/day) × Days] ÷ [Concentration (mg/L) × 10⁶]

For a 100 m³/h plant at 30 mg/L dose running 7 days, using 10% PAC (≈100,000 mg/L active): 30 × 2,400 × 7 ÷ 100,000 = ~5.0 m³ working volume, before any safety factor. Add 1.5–2× for the bulk tank itself, plus a day tank sized for 8–24 h of consumption with a level switch that interlocks the transfer pump.

Material selection is non-negotiable. Cross-linked polyethylene or FRP is standard. Never dilute PAC with water in the storage tank — it gels irreversibly and ruins the batch (PWTAG, S3, 2014-02, still the standing warning in 2026). Codes to reference on the datasheet: API 650 for atmospheric tank fabrication, NFPA 30 for flammable-liquid containment and bunding, and a secondary-containment volume at 110% of the largest tank. A typical packaged approach uses skid-mounted automatic chemical dosing systems that integrate the day tank, mixer, level instrumentation, and metering pump on a single frame — this is the configuration most procurement teams actually order.

Step 3 — Select the Metering Pump and Piping

Pump topology is decided by flow and turndown. Peristaltic pumps win at the low end — they hit the 0.1 ml/m³ accuracy band that PWTAG specifies for pool duty (S3) and have no seals to leak. At industrial flows the workhorse is a mechanical-diaphragm coagulant metering pump with a turndown of ≥100:1 so it can ride through the 2–3× flow swings that show up in real raw water. ANSI/HI 9.6.6 is the standard to cite for reciprocating positive-displacement metering-pump performance and turndown verification; ASME B73.1 covers chemical process pumps more generally and should appear on the data sheet as a baseline reference.

Sizing equation:

Pump capacity (L/h) = [Dose (mg/L) × Flow (m³/h) ÷ PAC concentration (mg/L)] × 1000

Apply a 1.5–2× safety factor, then check that minimum pump capacity at the lowest credible flow still exceeds the dose the chemistry actually needs. A 100 m³/h plant at 30 mg/L using 10% PAC needs 30 L/h of active solution; with a 1.5× factor, specify a 45 L/h pump that can turn down to 0.45 L/h for low-flow periods — well within a 100:1 turndown envelope.

Wetted parts should be PVC, PP, or PVDF; diaphragms in EPDM or Viton. Avoid carbon steel — it corrodes and contaminates the coagulant with iron, which downstream operators will see as a sudden colour excursion. Calibration cadence: factory calibration at commissioning, then in-situ volumetric verification quarterly and after every diaphragm replacement. The table below compares pool-duty hardware against industrial-duty hardware for the same chemical.

An integrated automatic chemical dosing system skid that pairs the storage, mixing, level instrumentation, and metering pump on one frame is the practical way to deliver this combination of hardware and control on a single purchase order.

Step 4 — Injection Hydraulics and Contact Time

Dose accuracy is wasted if the chemical never sees the water. Injection velocity into the main process flow should stay at or below 1.5 m/s so the jet does not shoot past the reaction zone (PWTAG, S3). Coagulation contact time needs to be at least 10 seconds in the pipe or flash-mix chamber, with an additional 30–60 seconds before any polymer is added if a flocculant aid follows (HydroChemix, S2). Sequence matters: PAC first, polymer second, or the polymer coats the colloids and blocks the coagulant from working.

Injection hardware is a quill or lance with a check valve to prevent backflow during pump shutdown. Place the sample point for the controller upstream of the injection point but downstream of any prior chemical addition — that way the controller sees the water it is about to dose, not the water it has already dosed. For open tanks or basins where the pipe velocity is unreliable, specify a static mixer or a pumped recirculation loop sized to deliver a velocity-gradient × time (G·t) ≥ 10⁴ before the clarifier feedwell. G·t below 10⁴ is the usual reason a "correctly metered" dose still produces pinpoint floc that won't settle.

Step 5 — Control Architecture and the ROI of Closed-Loop Dosing

Open-loop flow pacing from a magflow meter with the dose set from the last jar test is the baseline; it is also the most common configuration in 2026 and the easiest to defend in a procurement review. The upgrade path is a streaming current control loop at the clarifier inlet, trimming the dose setpoint in real time as raw-water charge demand shifts. A 2025 full-scale drinking-water study using an ANN–GA hybrid model reported a 49% reduction in median PAC-related treatment cost and an increase in color-limit compliance from 52% to 63%, while holding pH compliance above 97% (MDPI, S5, 2025-08). That 49% is a drinking-water number and should be treated as the upper bound; industrial buyers should pilot a streaming-current loop before scaling it plant-wide.

Minimum PLC I/O list for a defensible specification:

• Analog in (4–20 mA): flow from magflow on the raw-water line
• Analog in (4–20 mA): day-tank level
• Analog out (4–20 mA or 0–10 V): pump speed command, 0–100%
• Digital in: day-tank low-level interlock (pump inhibit on empty)
• Digital in: transfer-pump run confirmation
• Analog in (optional, closed loop): streaming-current sensor, 4–20 mA

The cost framing matters because it is the language procurement uses. For a 100 m³/h plant at 30 mg/L dose using 10% PAC solution, consumption is 30 kg/h of active product. A 10% overdose wastes ~3 kg/h continuously — that is roughly 26,000 kg/year of coagulant, which at typical industrial chemical prices pays for a streaming-current retrofit in under 12 months on chemical savings alone. This is the single number that converts a dosing-accuracy argument into a budget line item, and it is the link to broader sludge dewatering system design criteria and RO water purification design criteria where over- and under-dosing also show up as avoidable opex.

Note: payback figures are illustrative, based on a 100 m³/h plant at 30 mg/L dose with 10% PAC solution and industrial-grade coagulant pricing in 2026.

Frequently Asked Questions

What is the typical PAC dose range for industrial wastewater?

Across most industrial wastewaters, commercial polyaluminum chloride is dosed at 5–150 mg/L as product, with an effective pH window of 5.5–8.5 (RubMech, S4). In low-alkalinity drinking water the practical operating band is much tighter at 4–12 ppm, per a 2025 GA-optimized full-scale study (MDPI, S5).

How do you size a PAC storage tank?

Use Tank volume (m³) = [Dose × Flow × Days] ÷ [Concentration × 10⁶], then apply a 1.5–2× safety factor. A 100 m³/h plant at 30 mg/L using 10% PAC needs roughly 5 m³ of working volume per week (S5 concentration). Always design secondary containment to 110% of the largest tank per NFPA 30, and never dilute PAC in the storage tank — it gels (PWTAG, S3).

Which metering pump standard applies to PAC dosing?

ANSI/HI 9.6.6 is the standard for reciprocating positive-displacement metering-pump turndown and performance verification, and should be cited on the data sheet. ASME B73.1 covers chemical process pumps more generally. Specify a turndown of ≥100:1 for industrial service, with mechanical-diaphragm or peristaltic wetted paths in PVC, PP, or PVDF.

What controls do I need on a PAC dosing skid?

At minimum: analog in for flow (4–20 mA), analog in for day-tank level, analog out for pump speed (0–100%), and a digital low-level interlock. Adding a streaming-current sensor at the clarifier inlet for closed-loop trim can reduce PAC-related treatment cost by up to 49% in full-scale drinking-water service (MDPI, S5, 2025-08); industrial payback is typically under 12 months on chemical savings.

What happens if I overdose or underdose PAC?

Overdosing reverses the colloidal charge, producing a cloudy supernatant and increased sludge volume (HydroChemix, S2). Underdosing passes turbidity downstream and shortens filter runtime. Either direction drives operating cost up; a flow-paced or streaming-current-trimmed loop is the standard 2026 mitigation.